Three years after a deadly winter storm nearly imploded the ERCOT grid, killed hundreds of Texans and caused billions in financial damages after blackouts lasted for days, stakeholders in the Texas market have begun working on a reliability standard that may be stricter than industry norms.
ERCOT is proposing a “multi-metric” framework that establishes thresholds on three criteria: frequency, duration and magnitude of loss-of-load events.
Its baseline recommendations would set a loss-of-load expectation (LOLE) frequency of once every 10 years; 14 hours of rolling outages during an event; and no more than 19 GW of load shed to maintain the ability to roll the outages (54584).
The grid operator said using maximum magnitude as a probabilistic measure addresses a key physical reliability constraint: how many megawatts can be effectively managed at one time for rotating load shed purposes. It included maximum duration because one reliability policy constraint is the acceptable length to customers of an outage event.
Pete Warnken, ERCOT senior manager of resource adequacy, told the Texas Public Utility Commission during a May 2 technical workshop that after Winter Storm Uri in 2021, it became clear that the industry’s normal one-in-10 LOLE wasn’t enough on its own. He said staff reviewed other grid operators’ reliability standards and dug into background materials to come up with their proposal.
“One overarching theme became apparent: Simply relying on the 0.1 LOLE industry standard was not acceptable, and any reliability standard for ERCOT needed to expand beyond this single metric,” Warnken said. “There is an expectation for the commission to establish a reliability standard for ERCOT and take action to ensure the reliability and needs of the region are met both in the near and long term.”
The 2021 storm came 10 years after a less severe cold weather event in 2011. The rolling outages during the week leading up to Super Bowl XLV, played in the Dallas-Fort Worth area, were shorter and less severe than Uri’s.
“It makes me think that at a basic level, we are hitting that one-in-10 standard, but we’re still getting the massive outages that we want to try to avoid,” Commissioner Jimmy Glotfelty said. “So, semantics. Two massive outages in 20 years, that’s one in 10.”
The commission and stakeholders generally supported ERCOT’s approach.
“I think what ERCOT is proposing makes sense,” PUC Chair Thomas Gleeson said, expressing more interest in what market participants had to say.
“This is probably the most important policy decision this commission is going to make in terms of the impact to the state and reliability for our system,” NRG Energy’s Bill Barnes said, adding that his company “strongly supports” the resource adequacy-based reliability standard.
“We feel that this is the missing piece of our market structure. For the most important reliability type of our grid, resource adequacy, up to this point it’s been a shoulder shrug and, ‘Let’s just see what we get.’ That’s why this is such an important decision,” he added.
Katie Coleman, representing Texas Industrial Energy Consumers and its large industrial users, said the standard could be a “useful tool” as a reference point to decisions on whether to increase the offer cap, change the shape of the operating reserve demand curve or add ancillary services.
“There’s a lot of judgment involved in a reliability standard. It’s extremely imprecise,” Coleman said. “We continue to have concerns about using it as a single reference point to move billions of dollars around through a capacity construct. So that’s our sensitivity, but not the reliability standard in and of itself.”
‘Reasonable Starting Point’
PUC staff have since filed a memo responding to several points made during the technical conference. It lays out the decision points staff say it needs to prepare a proposal for the reliability standard’s rulemaking.
The commissioners will use the memo as the basis for discussions during their May 16 and 23 open meetings. A final rule could possibly be published by June 13, and a final PUC vote taken on the rule in August.
Commission staff said they view ERCOT’s approach to a reliability standard recommendation to be a “reasonable starting point” and that a commission-approved standard is “essential to achieving long-term resource adequacy.” They said setting the LOLE at close to one event every 10 years is a “reasonable benchmark” that alternative values can be compared to.
“At a minimum, the commission-approved reliability standard should target a level of reliability that is comparable to other markets and regions across the country,” they said in the memo.
Staff also noted that adopting a reliability standard does not require implementing the performance credit mechanism (PCM), saying it is not the only tool that could be used to meet the standard. They suggested “alterations” to existing ancillary service products, new reliability products or changes to the scarcity pricing signals as other policy options that could be “tailored” to affect reliability standard metrics.
While staff agreed with using the industry’s one-in-10 LOLE standard, they found setting a firm megawatt value for the 19-GW magnitude metric is not appropriate as it is directly tied to the system’s operational capability. They suggested a 0.25% exceedance probability for magnitude and updating the metric on a predictable, scheduled basis that aligns with future load-shed capabilities.
Staff also recommended the duration metric be reduced to 12 hours, with a “more relaxed” 1% exceedance probability. They noted ERCOT’s emergency pricing program will kick in after prices have been at the high systemwide offer cap for more than 12 consecutive hours.
According to ERCOT’s cost analysis, a 0.1 LOLE is not enough to constrain the maximum magnitude to 19 GW; instead, it would require a 0.04 LOLE. The incremental system cost to achieve this increased reliability is between $195 million and $271 million per year above the amount that supports a 0.1 LOLE, staff said.
ERCOT’s sensitivity variables include using weather years dating back to 1980 to ensure a “robust weather history” is accounted for. It also suggests a retirement assumption of 900 MW over the next several years and using combustion turbines for capacity, as the latter can be converted into any other combination of resource types.